Model selection configuration for EMT data.

For model selection at the sparse regression step, we considered basis libraries that include polynomial terms and/or Hill terms — poly_max_2: polynomial terms of degree up to 2 (bias term included); poly_1_2: polynomial terms of degrees 1 and 2 (no bias term); hill_n_poly_1: Hill terms of order n plus linear terms; hill_n_poly_xy: Hill terms of order n plus linear terms and interaction terms of the form xy; hill_n_poly_1_2: Hill terms of order n plus polynomial terms of degrees 1 and 2; and hill_max_3_poly_1: Hill terms of order up to 3 plus linear terms. (XLSX)